Skip to content
conftrace_
2021 SemEval SemEval 2021

LIIR at SemEval-2021 task 6: Detection of Persuasion Techniques In Texts and Images using CLIP features

Abstract

AbstractWe describe our approach for SemEval-2021 task 6 on detection of persuasion techniques in multimodal content (memes). Our system combines pretrained multimodal models (CLIP) and chained classifiers. Also, we propose to enrich the data by a data augmentation technique. Our submission achieves a rank of 8/16 in terms of F1-micro and 9/16 with F1-macro on the test set.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Natural Language Processing
🧭 Keyword Pioneer - clip model
🐣 Hot Topic Early Bird - clip model

Authors

Erfan Ghadery, Damien Sileo, Marie-Francine Moens

Topics

Artificial Intelligence > Core AI > Multimodal Learning Machine Learning > Core Methods > Classification Machine Learning > Application Areas > Data Augmentation Natural Language Processing > Applications > Text Classification

Keywords

image classification transfer learning feature extraction text classification data augmentation multimodal learning multi-label classification persuasion detection clip model
Download PDF

More from SemEval 2021

All SemEval 2021 papers →
UTNLP at SemEval-2021 Task 5: A Comparative Analysis of Toxic Span Detection using Attention-based, Named Entity Recognition, and Ensemble Models
macech at SemEval-2021 Task 5: Toxic Spans Detection
LZ1904 at SemEval-2021 Task 5: Bi-LSTM-CRF for Toxic Span Detection using Pretrained Word Embedding
AIMH at SemEval-2021 Task 6: Multimodal Classification Using an Ensemble of Transformer Models
HOMADOS at SemEval-2021 Task 6: Multi-Task Learning for Propaganda Detection
1213Li at SemEval-2021 Task 6: Detection of Propaganda with Multi-modal Attention and Pre-trained Models